Siyuan Li, Jing Li, Neil Armstrong, Junxi Wang, Ruitong Liu, Youyuan Zhang
We lifted 9 functions out of this paper's own repositories and ran 4 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| lab-klc/ScopeEdit | canonical | 4 of 9 |
| Function | Status | Where it lives |
|---|---|---|
| binary_log_probs | Ran | lab-klc/ScopeEdit/easyeditor/trainer/losses.py code served (permissive licence) · get_code("27659a9c234ffb11") |
| disabled_train | Ran | lab-klc/ScopeEdit/easyeditor/trainer/blip2_models/blip2.py code served (permissive licence) · get_code("4cb732f513d69dfd") |
| masked_mean | Ran | lab-klc/ScopeEdit/easyeditor/trainer/losses.py code served (permissive licence) · get_code("e078eeed20f03838") |
| tile | Ran | lab-klc/ScopeEdit/easyeditor/trainer/blip2_models/base_model.py code served (permissive licence) · get_code("cb33571427334815") |
| all_gather_with_grad | Not yet run | lab-klc/ScopeEdit/easyeditor/trainer/blip2_models/base_model.py code served (permissive licence) · get_code("0ec9fc2025c16f65") |
| compute_sim_matrix | Not yet run | lab-klc/ScopeEdit/easyeditor/trainer/blip2_models/blip2.py code served (permissive licence) · get_code("cdbd3e50b9bc4a2b") |
| concat_all_gather | Not yet run | lab-klc/ScopeEdit/easyeditor/trainer/blip2_models/base_model.py code served (permissive licence) · get_code("a829872cf2507355") |
| get_model | Not yet run | lab-klc/ScopeEdit/easyeditor/trainer/models.py code served (permissive licence) · get_code("ed9a3009555975a1") |
| kl_loc_loss | Not yet run | lab-klc/ScopeEdit/easyeditor/trainer/losses.py code served (permissive licence) · get_code("bcc592aa5c1e2bb3") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Online multimodal knowledge editing requires injecting a continual stream of visual-textual corrections into multimodal large language models (MLLMs) with bounded overhead and minimal disruption to unrelated behaviors. Existing editors mainly emphasize edit reliability and long-horizon stability, but rarely control the semantic boundary of each edit. Our pilot analyses of post-edit behaviors and internal neuronal activities reveal a scope gap behind reliable edits: instance-level success neither guarantees transfer to valid cross-modal variants nor prevents leakage to unrelated inputs, while edit-related cross-modal responses concentrate in deeper semantic layers. Therefore, we formulate Edit-Scoped Generalization, reframing online MLLM editing from merely correcting an instance to controlling the propagation boundary of each edit. To this end, we propose ScopeEdit, a scope-aware online editor that decomposes each update into a modality-local absorption branch and an evidence-gated shared generalization branch. The local branch supports stable edit absorption, whereas the shared branch enables cross-modal propagation only when visual and textual evidence are sufficiently aligned. Both branches perform scope-separated write geometries in orthogonal low-rank spaces and maintain branch-wise preconditioners via Sherman--Morrison recursions, yielding constant per-edit overhead. Extensive experiments across diverse benchmarks, long-horizon edit streams, MLLM backbones, real-world VLKEB scenarios, and complex vision-language architectures show that ScopeEdit consistently improves the trade-off between in-scope cross-modal transfer and out-of-scope locality, while preserving edit reliability, stability and online efficiency. Our code is available at https://github.com/lab-klc/ScopeEdit.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2607.01978")
get_code_for_paper("2607.01978")
have("2607.01978")
Connect an agent — have() is free.